Geo News
Community curated by people like you
LatestAICryptoHealthWorld AffairsUS Politics

Hardware manufacturing stories

Aug 24, 2026

Nvidia begins mass production of Groq 3 LPX AI inference chips

Nvidia has put its Groq 3 LPX racks into full production following its $20 billion acquisition of Groq, with the SRAM-based language processing units achieving 3,400 tokens per second and Samsung manufacturing the 4-nanometer chips. The first deployment is set for Nebius Token Factory before year-end.

Aug 24, 2026·4 sources
00
Aug 12, 2026

Foxconn reports 35% profit surge on AI hardware demand

Taiwan's Foxconn reported a 35% increase in second-quarter profit, beating analyst forecasts, driven by continued strong global demand for AI hardware and servers.

Aug 12, 2026·3 sources
00
Aug 5, 2026

Foxconn reports record monthly revenue of $29.3 billion in July, driven by AI product demand

Taiwanese electronics manufacturer Foxconn (Hon Hai Precision Industry Co.) posted record consolidated revenue exceeding NT$900 billion ($29.3 billion) in July 2026, marking a 54% year-over-year increase attributed to surging demand for artificial intelligence products.

Aug 5, 2026·8 sources
00
Jul 9, 2026

Meta to Begin Manufacturing AI Chip in September, Plans to Double Computing Capacity

Meta Platforms plans to start manufacturing its own artificial intelligence chip from September as part of a strategy to boost overall computing power to 14 gigawatts next year, according to an internal memo reviewed by Reuters.

Jul 9, 2026·1 source
00
Jul 6, 2026

Nvidia's Kyber NVL144 AI Rack System Delayed to 2028 Due to Manufacturing Issues, Asian Supplier Stocks Drop

Nvidia's next-generation Kyber NVL144 rack-scale system, designed to house Rubin Ultra chips, has been delayed by over 12 months to 2028 due to circuit board manufacturing problems, according to analyst firm SemiAnalysis. The delay sent Asian technology supplier stocks tumbling, with some losing double-digit percentages in market value.

Jul 6, 2026·9 sources
00

Top claims

  • ▪The Groq 3 LPX system achieved 3,400 output tokens per second running Gemma 4 31B in benchmarks conducted by Artificial Analysis.
  • ▪Nvidia claims the Groq 3 LPX delivers a 4x reduction in decode latency compared to the nearest GPU-based alternative.
  • ▪The Groq 3 LPU stores model weights in on-chip static random-access memory (SRAM) to deliver 150 terabytes per second of bandwidth per chip.

Subtopics

Compute, chips & AI infrastructure5AI2Corporate earnings2AI Data Center Gold Rush1AI inference (scaling)1AI startups1AI supply chain1AI tokens1Data platforms1Semiconductor manufacturing1

Related timelines

Payments

118 stories

Trump administration

116 stories

Congress

108 stories

Iran War

209 stories

Russia-Ukraine war

136 stories